DescriptionThis dissertation includes four essays on empirical asset pricing as well as the application of state-space models in this area.
The first essay seeks to reconcile the debate about the price effect of risk-neutral skewness (RNS) on stocks. We document positive predictability from short-term skewness, consistent with informed trading, and negative predictability from long-term skewness, consistent with skewness preference. A term spread on RNS captures different information from long- and short-term contracts, resulting in stronger predictability. The quintile portfolio with the lowest spread outperforms that with highest spread by 14.64% annually. The information difference between short- and long-term options explains the pricing difference of their RNS, providing a potential resolution to the debate.
The second essay uses a novel sequential Monte Carlo method, the mixture Kalman filter (MKF), to detect periodically collapsing rational bubbles in stock prices. The stock-dividend-bubble system is expressed in a state-space model with Markov regime switching. We apply the MKF to estimate the model for simulated and actual stock data. Our methodology captures the bubble dynamics more successfully than the model without regime-switching, and identifies major bubble collapsing episodes in our sample period.
The third essay explains why the correlation between oil price and 10-year TIPS break-even inflation increased dramatically after the financial crisis. We develop a shadow-rate no-arbitrage term structure model to fit nominal yields, TIPS yields and inflation forecasts, and estimate it using the Extended Kalman Filter. Based on the model estimation, we provide empirical evidences showing that the puzzle is because when interest rates bind at the zero lower bound, investors doubt the effectiveness of monetary policy to control deflation. We justify this mechanism theoretically under a general equilibrium framework.
The last essay develops an arbitrage-free Nelson-Siegel term structure model with economic factors, estimates the model using the Kalman Filter and reveals the information content of treasury term premium through model decomposition. We find foreign carry trades and cross-border investments have been the key drivers of negative treasury term premium in recent years due to monetary divergence. This poses challenges to conventional term structure models using US-exclusive macro variables to explain treasury dynamics.